The effects of repeated exposure to hypercapnia on arousal and cardiorespiratory responses during sleep in lambs

Authors

  • Renea V. Johnston,

    1. Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Monash University, Clayton, Victoria, 3168, Australia
    Search for more papers by this author
  • Daniel A. Grant,

    1. Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Monash University, Clayton, Victoria, 3168, Australia
    Search for more papers by this author
  • Malcolm H. Wilkinson,

    1. Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Monash University, Clayton, Victoria, 3168, Australia
    Search for more papers by this author
  • Adrian M. Walker

    1. Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Monash University, Clayton, Victoria, 3168, Australia
    Search for more papers by this author

Corresponding author A. M. Walker: Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Level 5, Monash Medical Centre, 246 Clayton Road, Clayton, Victoria 3168, Australia. Email: adrian.walker@med.monash.edu.au

Abstract

Arousal and cardio-respiratory responses to respiratory stimuli during sleep are important protective mechanisms that rapidly become depressed in the active sleep state when episodes of hypoxia or asphyxia are repeated: whether responses to repeated hypercapnia are similarly depressed is not known. This study aimed to determine if arousal and cardio-respiratory responses also become depressed with repeated episodes of hypercapnia during sleep and whether responses differ in active sleep and quiet sleep. Eight newborn lambs were instrumented to record sleep state and cardio-respiratory variables. Lambs were subjected to two successive 12 h sleep recordings, assigned as either sequential control and test days, or test and control days performed between 12.00 and 00.00 h. The control day was a baseline study in which the lambs breathed air to determine spontaneous arousal probability. During the test day, lambs were exposed to a 60 s episode of normoxic hypercapnia (Fractional inspired CO2(Finline image)= 0.08 and Fractional inspired O2(Finline image)= 0.21 in N2) during every quiet sleep and active sleep epoch. The probability of lambs arousing during the hypercapnic exposure exceeded the probability of spontaneous arousal during quiet sleep (58%versus 21%, χ2= 54.0, P < 0.001) and active sleep (39%versus 20%, χ2= 10.0, P < 0.01), though the response was less in active sleep. Exposure to hypercapnia also resulted in a significant increase in ventilation in quiet sleep (150 ± 22%) and active sleep (97 ± 23%, P < 0.05), though the increase was smaller in active sleep (P < 0.05). Small (< 5%) blood pressure increases and heart rate decreases were evident during hypercapnia in quiet sleep, but not in active sleep. Arousal and cardio-respiratory responses persisted with repetition of the hypercapnic exposure. In summary, although arousal and cardio-respiratory responses to hypercapnia are less in active sleep compared with quiet sleep, these protective responses are not diminished with repeated exposure to hypercapnia.

Arousal from sleep and cardio-respiratory compensations are powerful protective responses to respiratory stresses such as asphyxia (Campbell et al. 1998), asphyxia associated with airway occlusion (Sullivan et al. 1978; Igras & Fewell, 1991; Davidson & Fewell, 1994; Harding et al. 1997), and the components of asphyxia, hypoxia (Phillipson et al. 1978; Henderson-Smart & Read, 1979; Jeffery & Read, 1980; Neubauer et al. 1981; Berthon-Jones & Sullivan, 1982; Fewell & Baker, 1987) and hypercapnia (Phillipson et al. 1977; Douglas et al. 1982; Berthon-Jones & Sullivan, 1984; Fewell & Baker, 1989). However, indirect evidence suggests that protective arousal responses may become ineffective if the respiratory stress is repeated in sleep. For example, adults with obstructive sleep apnoea (OSA) terminated by arousal experience lengthening of the apnoea episode through the night (Charbonneau et al. 1994), and infants who experience repeated episodes of apnoea during sleep may succumb to Sudden Infant Death Syndrome (SIDS) (Kahn et al. 1992).

In experimental studies, asphyxia induced by airway obstruction is a potent arousing stimulus, but it becomes less effective with repetition (Fewell et al. 1988; Harding et al. 1997; Brooks et al. 1997). Arousal also becomes depressed after repeated exposure to one of the components of airway obstruction-induced asphyxia, namely hypoxia (Fewell & Konduri, 1989; Johnston et al. 1998). Concurrent with the depression of arousal during repetitive hypoxia is a loss of the ventilatory and blood pressure augmentation that normally occurs in response to hypoxia (Johnston et al. 1999). It has been suggested that when the arousal response to hypoxia becomes depressed, hypercapnia may assume importance for initiating arousal from sleep (Fewell & Konduri, 1988). Thus, the question of whether arousal and cardio-respiratory responses to hypercapnia, like those to hypoxia, also become impaired with repetition of the stimulus is critical to understanding increased vulnerability to repeated asphyxial stress. However, no study has as yet addressed the issue of arousability and cardio-respiratory responses to repeated hypercapnic stress.

Whether responses to repeated hypercapnia are modulated by sleep state is also unknown. Some studies have demonstrated that the depressant effects upon arousal of hypoxia (Fewell & Konduri, 1989) and asphyxia (Brooks et al. 1997) occur in both rapid-eye-movement sleep (REM sleep, or active sleep as it is termed in the newborn) and also in non-REM (termed quiet sleep in the newborn). Others have shown that arousal depression during hypoxia (Johnston et al. 1998) and airway obstruction (Fewell et al. 1988; Harding et al. 1997) is confined to the REM phase. As OSA occurs predominantly during active sleep in the human newborn (Kahn et al. 1992) and the de-saturations associated with OSA are more severe during this state in the adult (Sullivan & Issa, 1980; Charbonneau et al. 1994), active sleep may be the more vulnerable to arousal and cardio-respiratory depression during repetitive asphyxial stress.

By examining cardio-respiratory and arousal responses during a sequence of repeated exposures to hypercapnia during sleep in newborn lambs, this study aimed to establish whether arousal and cardio-respiratory responses are depressed by repetitive hypercapnia as they are with exposure to repetitive hypoxia. In addition, the study aimed to determine if there are differences between active sleep and quiet sleep in responses to repetitive hypercapnia.

Methods

Lambs were chronically implanted with instrumentation to measure sleep and wakefulness and cardio-respiratory parameters by methods previously described (Johnston et al. 1998, 1999). All surgical and experimental protocols were performed in accordance with the guidelines established by the National Health and Medical Research Council of Australia and with the approval of the Standing Committee in Ethics in Animal Experimentation of Monash University. At the completion of studies, lambs were killed with a lethal dose of anaesthestic (150 mg kg−1 sodium pentobarbitone, intravenously).

Animals

Eight newborn Border–Leicester × Merino cross lambs (body mass 4.2 ± 0.2 kg, mean ±s.e.m.) were delivered spontaneously at term (147 days) on the premises. Lambs were separated from their ewes 12–48 h after birth, housed with another lamb and taught to feed unaided prior to surgery (Lamb Milk Replacer, Veanavite, Shepparton, Australia). Lambs were weighed every 1–2 days to ensure adequate weight gain (on average they gained 221 ± 12 g day−1) and rectal temperature was measured every 2–3 days to ensure they were generally well and without infection, both prior to and after surgical intervention.

Surgical preparation

Sterile instrumentation was surgically implanted in lambs (3–11 days old) to measure cerebral electrical activity and certain cardio-respiratory parameters. Anaesthesia was induced with 2% halothane in 60% oxygen (and balance nitrous oxide) delivered via a face mask, then lambs were intubated with an endotracheal tube (i.d. 5.0 mm, Mallinckrodt Medical, Athlone, Ireland) and mechanically ventilated (1–1.5% halothane in 60% oxygen, balance nitrous oxide). An incision was made at the right axilla and vascular catheters (Tygon, o.d. 1.5 mm) were chronically implanted into the axillary artery and axillary vein using a non-occlusive technique for measurement of blood pressure and blood sampling. A right lateral thoracotomy was performed at the sixth intercostal space and a liquid-filled balloon-tipped catheter (volume: 1–2 ml) was inserted into the intrapleural cavity to record changes in intrapleural pressure (Ppl). The amplitude of the deflections in Ppl provided an index of inspiratory and expiratory effort (Harding et al. 1997). For sleep state definition, paired Teflon-coated stainless steel electrodes (Medwire, New York) were implanted on the parietal cortex to record the electrocorticogram (ECoG), at the inner and outer canthus of the right eye to record the electro-oculogram (EOG) and in the dorsal musculature of the neck to record the electromyogram (EMGn). All electrodes were referenced to a single electrode sewn into the subcutaneous tissue of the scalp, and the leads were tunnelled subcutaneously and exited the skin at the back of the neck approximately 2–3 cm posterior to the EMGn incision site. A tracheostomy was performed, the endotracheal tube was removed, and a fenestrated tracheostomy tube (i.d. 5.0 mm; Shiley Inc., Irvine, CA, USA) implanted in the trachea to allow rapid changes in inspired gas mixtures. The tracheostomy tube was capped to allow the lamb to breathe normally via the upper airway, or during experimental procedures, an inner cannula was inserted in the tracheostomy tube so that the lamb breathed via an external circuit.

Physiological measurements

Lambs were allowed a minimum of 72 h recovery from surgery before they were studied in controlled temperature conditions (22–25°C). The lamb cage was partitioned during recordings to allow the lamb enough space to stand up, lie down and feed freely, while restricting it from turning around. The arterial and the intrapleural catheters were connected to calibrated strain gauge manometers (Cobe CDX III, Cobe Laboratories, Lakewood, CO, USA) and referenced to the mid-thoracic level when the animal was lying down. A pulse oximeter probe was placed around the tail to continuously measure arterial oxygen saturation (Sinline image, Model N200, Nellcor Inc., Hayward, CA, USA). End-tidal CO2 (Pinline image; Engstrom Eliza, Gambro, Melbourne, Australia) was measured in the tracheostomy tube. The strain gauge manometer, the pulse oximeter, the CO2 analyser and the electrodes were connected to a signal conditioner (Cyberamp 380, Axon Instruments, Foster City, CA, USA). Arterial pressure, pleural pressure, Sinline image and Pinline image were low-pass filtered at 100 Hz. The electrophysiological signals were filtered at 30–100 Hz for the EMGn, and 0.3–40 Hz for the EOG and the ECoG. All signals were continuously displayed on a thermal chart recorder (Model 7758A, Hewlett-Packard, Waltham, MA, USA). Data were also intermittently stored on a personal computer (486 DX/50) at a sampling rate of 200 Hz, using an analog–digital converting board (ADAC 4801A, ADAC, Woburn, MA, USA) and analysis software (CVSOFT Data Acquisition and Analysis Software, Odessa Computer Systems, Calgary, Canada).

Lambs were subjected to two successive 12 h sleep recordings, separated by a 12 h interval, assigned as either sequential control and test days, or test and control days, and conducted between 12.00 and 00.00 hours. Duplicate studies were conducted in two of eight lambs, after allowing 6–7 days recovery, for a total of 10 studies over an age range of 7–30 days (14 ± 2 days, mean ±s.e.m.).

During the control day of the study, lambs were assigned to breathe room air via a breathing circuit. During the test day of the study, the lambs were exposed to repeated challenges of a gas mixture designed to produce normoxic hypercapnia (Finline image= 0.08 and Finline image= 0.21 in N2, via the breathing circuit) during every active sleep and quiet sleep epoch.

Quiet sleep was defined by the presence of high-voltage, low-frequency waves on the ECoG, absence of eye movements and reduced EMGn activity compared with the awake state. Arousal from this state was characterized by a change in the ECoG to low-voltage, high-frequency waves, increased EMGn activity, and opening of the eyes (Fig. 1). Active sleep was defined by the presence of low-voltage, high-frequency waves on the ECoG, the presence of rapid eye movements and the absence of EMGn tone. Arousal was characterized by a return of tonic activity in the EMGn (Fig. 1).

Figure 1.

Arousal and cardio-respiratory responses during hypercapnia in sleep
Physiological recording in a sleeping lamb, illustrating cardio-respiratory and arousal responses to acute hypercapnia during a period of quiet sleep (QS, left) and active sleep (AS, right). Arousal to the waking state (W) is signified by the vertical dashed line. The 60 s hypercapnia stimulus, induced by rapidly changing the air in the inspiratory line to a gas mixture of Finline image= 0.21 and Finline image= 0.08 in N2, is denoted by the horizontal bar. BP, blood pressure, Ppl, pleural pressure, Pinline image, end-tidal CO2; EMGn, nuchal electromyogram; EOG, electro-oculogram; ECoG, electrocorticogram.

Arterial blood was sampled daily while the catheter was opened during flushing, and prior to the commencement of the control, test and recovery periods of each study, to ensure that blood gases and pH were within normal limits for lambs in this laboratory. Lambs were awake while blood was sampled during these times. Additional samples were obtained during periods of sleep during the 12 h test period of the study. Arterial blood was withdrawn during random periods of air-breathing and during periods of hypercapnia, for both sleep states. Blood (0.4 ml) was collected in heparinized 1 ml syringes for analysis of oxygen saturation (SaO2, OSM2 Hemoximeter, Radiometer Pacific, Copenhagen, Denmark), and pH and blood gases corrected to 39.5°C (ABL500, Radiometer Pacific, Copenhagen, Denmark).

Analysis of sleep and arousal data

Sleep patterns were quantified by determining the number and duration of sleep epochs, and total sleep time in quiet sleep and active sleep for the control and test days. A total of 249 quiet sleep epochs and 283 active sleep epochs on the control day, and 161 quiet sleep epochs and 141 active sleep epochs on the test day were available for analysis. Paired t tests were used to detect differences in the number of sleep epochs, epoch length, and total sleep time across the two study nights. Probability analysis was used to quantify the arousing effect of the exposures to hypercapnia. Probability analysis was limited to the first 40 epochs on each of the two study nights; too few animals were subjected to a greater number of exposures to hypercapnia to undertake probability analysis further. The probability of arousal to hypercapnia in each sleep state was calculated as the percentage of tests in which arousal occurred during the 60 s stimulus. The probability of spontaneous arousal was determined on the control night; spontaneous probability was the percentage of arousals that occurred between 30 and 90 s of each epoch, matching the period in which hypercapnia tests were presented. Differences between the probability of arousal in hypercapnia and the probability of spontaneous arousal were tested using the χ2 test; these tests were performed separately for each of active sleep and quiet sleep, and for each of the sequential groups of 10 hypercapnia exposures. A χ2 test was also used to test for differences in probability of arousal in response to hypercapnia between sleep states. All probability data were stratified (Horne et al. 1989, 1991) to prevent bias arising from differences in the number of sleep epochs between animals.

A two-way ANOVA for repeated measures was used to determine if differences in blood gas and acid–base status existed between sleep states and between periods of breathing air and breathing the hypercapnic gas mixture.

Analysis of cardio-respiratory responses during repetitive hypercapnia

Cardio-respiratory variables: blood pressure, heart rate, Pinline image, respiratory amplitude (assessed from pleural pressure changes), respiratory frequency and an index of ventilation (the product of amplitude and frequency) were examined for every exposure to hypercapnia during the test day and analysed for n= 5 lambs (six studies). A total of 78 exposures to hypercapnia were employed for analysis in quiet sleep and 79 in active sleep. For this analysis, cardio-respiratory variables were measured for 10 breaths prior to arousal and for those tests that did not result in an arousal, over the 10 breaths that occurred just prior to the 30 s point of the hypercapnia exposure (a time point which corresponded to the average arousal time). Cardio-respiratory variables were pooled for arousal and non-arousal epochs, and as uneven numbers of tests were contributed by different lambs, a mean response was first calculated for each lamb, then an overall mean response was derived from the five individual responses. Paired t tests were used to detect differences in cardio-respiratory variables between periods of air and hypercapnia, and between sleep states. A one-way ANOVA for repeated measures was used to assess the effects of repetitive hypercapnia upon cardio-respiratory variables.

Results

Arterial blood gas analysis performed during sleep on the test night of the study confirmed that Pinline image was significantly higher, and pH was significantly lower during periods of breathing hypercapnic gas in both sleep states (P < 0.05, Table 1) compared with periods of air-breathing normocapnia. Pinline image also increased during periods of hypercapnia (P < 0.05) compared with air-breathing in quiet sleep, but these changes did not result in significant changes in Sinline image. There were no differences in base excess during periods of air-breathing and hypercapnia. There were no differences in blood gases or pH during quiet sleep and active sleep during air-breathing (Table 1).

Table 1.  Effects of hypercapnia on blood gases and pH during sleep in lambs
 Quiet sleepActive sleep
AirHypercapniaAirHypercapnia
  1. Values are mean ±s.e.m., n= 5 lambs. Sinline image, arterial oxygen saturation; Pinline image, arterial partial pressure of oxygen; Pinline image, arterial partial pressure of carbon dioxide. *P < 0.05, air versus hypercapnia (2-way ANOVA for repeated measures).

S inline image (mmHg)94 ± 2 95 ± 1 93 ± 194 ± 1 
P inline image (mmHg)101 ± 3  118 ± 3* 102 ± 5 111 ± 9  
P inline image (mmHg)42 ± 1 59 ± 1*42 ± 159 ± 1*
pH7.40 ± 0.02 7.33 ± 0.01* 7.41 ± 0.01 7.33 ± 0.01*
Base excess (mmol l−1)2.3 ± 0.63.2 ± 0.9 2.4 ± 0.73.5 ± 0.8

When examined over the entire 40 exposures to hypercapnia, continuous measurement of Pinline image confirmed the results of blood gas analysis of single samples, revealing that for both sleep states, the gas mixture applied resulted in a significant increase in Pinline image from air to hypercapnia. There were no differences in the level of hypercapnia (Table 2) in epochs which resulted in an arousal response (measured at the point of arousal) and those in which there was no arousal (measured after 30 s of hypercapnia). In all epochs of the entire 40 exposures to hypercapnia that did not result in an arousal, Pinline image (measured after 60 s of hypercapnia) did not differ between states (Table 2).

Table 2.  P inline image during hypercapnia
 AirHypercapnia
ArousalNA (30 s)NA (60 s)
  1. P inline image values (mmHg) (n= 5 lambs) for the entire series of 40 exposures to hypercapnia; values were measured during the 30 s period prior to hypercapnia (air), at the point of arousal, or at 30 s and 60 s of hypercapnia during epochs that did not result in arousal (NA).

Quiet sleepRange37–4155–6654–6856–67
Mean ±s.e.m.39 ± 160 ± 262 ± 364 ± 3
Active sleepRange38–4257–7355–6059–64
Mean ±s.e.m.40 ± 163 ± 358 ± 162 ± 1

Arousal responses to repetitive hypercapnia

In quiet sleep and active sleep, the imposition of repeated exposures to hypercapnia did not affect the number of sleep epochs on the test day compared with the control day (Table 3). Epoch length and total sleep time were reduced on the test day (P < 0.05, Table 3).

Table 3.  Sleep characteristics
 Study nightQuiet sleepActive sleep
  1. Values are mean ±s.e.m., n= 5 lambs). *P < 0.05, control versus test night (paired t test).

Number of epochsControl27 ± 117 ± 2
Test26 ± 216 ± 1
Epoch length (min)Control 5.2 ± 0.2 3.7 ± 0.2
Test  4.0 ± 0.2*  3.0 ± 0.3*
Total sleep time (min)Control141 ± 8 64 ± 6
Test103 ± 8* 45 ± 4*

When examined over the entire series of 40 exposures to repetitive hypercapnia, the overall probability of arousing from quiet sleep (58%) was approximately three times greater than the overall probability of spontaneous arousal from quiet sleep (21%, χ2= 54.0, P < 0.05; Fig. 2). A similar analysis in active sleep showed that the overall probability of arousal in hypercapnia (39%) was significantly higher than the overall spontaneous arousal probability (20%, χ2= 10.0, P < 0.05; Fig. 2). Arousal probability in hypercapnia was significantly less in active sleep than quiet sleep (χ2= 13.1, P < 0.001, Fig. 2).

Figure 2.

Probability of arousal from sleep during hypercapnia
Probability of arousal from sleep during a series of 40 exposures to hypercapnia (filled bars) versus the probability of spontaneous arousal (open bars) during quiet sleep and active sleep. Note that the probability of arousal from both quiet sleep and active sleep was significantly higher than the probability of spontaneous arousal in both states. *P < 0.05, probability of arousal during hypercapnia versus probability of spontaneous arousal (χ2= 54.0 in quiet sleep, χ2= 10.0 in active sleep). †P < 0.001, probability of arousal during hypercapnia in quiet sleep versus active sleep (χ2= 13.1).

When partitioned into sequential groups of 10 hypercapnia exposures, the probability of arousal during hypercapnia in quiet sleep remained much higher than the probability of spontaneous arousal throughout the sequence of hypercapnia tests (P < 0.05, Fig. 3). With just one exception, there were no differences in the probability of arousal in hypercapnia across the sequential groups of 10 exposures; arousal in hypercapnia during the first 10 hypercapnia tests was significantly higher than the probability of arousal in hypercapnia during exposures 31–40 (χ2= 5.5, P < 0.05) but all other groups were similar. When partitioned into sequential groups of 10 exposures to hypercapnia, the probability of arousal during hypercapnia in active sleep was significantly greater than the probability of spontaneous arousal only in exposures 21–30 (P < 0.05, Fig. 3). Nevertheless, arousal probability in hypercapnia was numerically higher throughout the series of exposures and similar across the sequential groups of 10 exposures (P > 0.05).

Figure 3.

Probability of arousal from sleep and corresponding ventilatory responses during repetitive hypercapnia
Probability of arousal from sleep during repetitive hypercapnia (filled bars) during quiet sleep (left upper panel) and active sleep (left lower panel) plotted against the number of exposures to hypercapnia, compared with the probability of spontaneous arousal during normocapnia (open bars). Associated ventilatory responses (mean ±s.e.m., n= 5) to repeated exposure to hypercapnia (measured as percentage change from normocapnia levels) are shown for quiet sleep (right upper panel) and active sleep (right lower panel). Note that there was no depression of arousal or ventilation with repeated exposure to hypercapnia. *P < 0.05, probability of arousal during hypercapnia versus probability of spontaneous arousal and ventilatory response to hypercapnia versus baseline ventilation during normocapnia; †P < 0.05, probability of arousal during hypercapnia versus probability of arousal during hypercapnia for the first 10 exposures to hypercapnia (χ2 tests).

Cardio-respiratory responses to repetitive hypercapnia

During quiet sleep, ventilation averaged over the entire series of 40 exposures to hypercapnia increased significantly from baseline ventilation measured during air-breathing (150 ± 22%, P < 0.005). In active sleep, ventilation also increased significantly during hypercapnia (97 ± 23%, P < 0.05), though the increase was smaller than in quiet sleep (P < 0.05). Partitioned into sequential groups of 10 exposures to hypercapnia, ventilation increased significantly in all groups compared with air during quiet sleep (P < 0.05, Table 4), but the increase during hypercapnia did not reach statistical significance in active sleep. There were no effects of repetition of hypercapnia; no differences existed in ventilation across the sequential groups of 10 exposures in either state (P > 0.05, Fig. 3).

Table 4.  Cardio-respiratory and P inline image responses during repetitive hypercapnia
 Quiet sleepActive sleep
1–1011–2021–3031–401–1011–2021–3031–40
  1. Values are mean ±s.e.m., n= 5 lambs. Values are expressed as a percentage change from normocapnia for the following: ΔV, change in ventilation; ΔA, change in respiratory amplitude; Δf, change in respiratory frequency; ΔBP, change in mean arterial blood pressure; ΔHR, change in heart rate. CO2, denotes actual Pinline image (mmHg) during hypercapnia. There is no depression of any cardio-respiratory variables with repetition of hypercapnia (one-way ANOVA for repeated measures).*P < 0.05 value versus normocapnia (paired t tests).

ΔV (%)140 ± 10*115 ± 29*208 ± 70*200 ± 56*111 ± 3583 ± 2788 ± 4485 ± 34
ΔA (%)74 ± 8* 61 ± 18*105 ± 2* 107 ± 23* 76 ± 2362 ± 2059 ± 2762 ± 26
Δf (%)35 ± 4*29 ± 5* 43 ± 15* 40 ± 13*15 ± 610 ± 2  8 ± 1011 ± 4 
ΔBP (%)  1 ± 0.42 ± 12 ± 12 + 1 3 ± 22 ± 22 ± 22 ± 3
ΔHR (%) –2 ± 0.5–3 ± 1 –4 ± 2 –4 ± 2 –2 ± 2–4 ± 2 –5 ± 1 –3 ± 2 
CO259 ± 3*62 ± 3*61 ± 1*60 ± 2* 58 ± 2*60 ± 2*59 ± 1*59 ± 2*

The components of ventilation, amplitude and frequency examined over the entire series of exposures to hypercapnia were significantly increased during hypercapnia compared with air in both quiet sleep (amplitude, 81 ± 11%, P < 0.01; frequency, 34 ± 5%, P < 0.01) and active sleep (amplitude, 67 ± 17%, P < 0.05; frequency, 12 ± 2%, P < 0.01). These responses partitioned into sequential groups of 10 exposures to hypercapnia are detailed in Table 4. In a similar fashion to the ventilatory data, amplitude and frequency increased significantly in all groups compared with air during quiet sleep (P < 0.05, Table 4), but the increases during hypercapnia did not reach statistical significance in active sleep. There were no effects of repetition of hypercapnia; no differences existed in ventilation across the sequential groups of 10 exposures in either state (P > 0.05, Table 4).

Overall, blood pressure during quiet sleep increased from 77 ± 2 mmHg during air to 78 ± 2 mmHg during hypercapnia (P < 0.05), and heart rate decreased from 188 ± 8 beats min−1 during air to 182 ± 7 beats min−1 during hypercapnia (P < 0.05). Blood pressure during active sleep did not change significantly (73 ± 1 mmHg during air to 75 ± 2 mmHg during hypercapnia), nor did heart rate (174 ± 8 beats min−1 during air, 169 ± 7 beats min−1 during hypercapnia). When examined over sequential groups of 10 exposures to hypercapnia, blood pressure and heart rate did not increase significantly from baseline air values in quiet sleep or active sleep, nor were there any significant differences in blood pressure or heart rate across the series of 10 exposures (Table 4) in either quiet sleep or active sleep. There were no differences in the hypercapnia stimulus across the groups of 10 exposures; Pinline image measurements did not differ with repeated exposure (Table 4).

Discussion

This study has revealed that arousal and cardio-respiratory responses to hypercapnia do not become depressed with repetition of the challenge in either quiet sleep or active sleep. This is in marked contrast to the rapid depression of arousal and cardio-respiratory responses with repeated exposure to hypoxia in active sleep (Johnston et al. 1998, 1999).

Arousal responses to repetitive hypercapnia

In contrast to hypoxia (Fewell & Konduri, 1989; Johnston et al. 1998), hypercapnic hypoxia (Waters & Tinworth, 2005) and upper airway obstruction (Fewell et al. 1988; Harding et al. 1997), hypercapnia per se appears not to become less effective as an arousing stimulus with repetition. The probability of arousal during hypercapnia was significantly greater than the probability of spontaneous arousal in both quiet sleep and active sleep. Over the course of the repeated exposures to hypercapnia, the probability of arousing ranged from 71 to 46% in quiet sleep; these probabilities were all significantly greater than the probability of spontaneous arousal (∼20%), signifying that hypercapnia remained a powerful arousing stimulus with repetition. In active sleep the arousing effects of hypercapnia were less than in quiet sleep, but again we found no evidence for it becoming less effective as an arousing stimulus with repetition.

We did find evidence of limited habituation, as arousability reached its lowest point (41% in quiet sleep and 26% in active sleep) after 40 exposures. Thus, responses may have declined further if the stimulus were to have been continued. Nevertheless, the pattern of responses to repetitive hypercapnia remained substantially different than those to repetitive hypoxia, where depression occurs rapidly, after just 10 exposures (Johnston et al. 1998). Rapid depression of the arousal response to hypoxia, along with relative preservation of the response to hypercapnia, mirrors the effects upon ventilatory responsiveness after prolonged hypoxia (Long et al. 1994). While responses to hypoxia are persistently abolished following 1 h of exposure to hypoxia, responses to hypercapnia are preserved, an effect ascribed to selective depression of neurons generating the response to hypoxia, perhaps by γ-aminobutyric acid (GABA) (Weil, 1994).

This study adds to previous animal studies showing that arousability is less in active sleep compared with quiet sleep. Other animal studies have found delayed arousal in active sleep in response to hypercapnia (dogs, Phillipson et al. 1977; Issa & Bitner, 1993; lambs, Fewell & Baker, 1989). Also arousal responses are slower in REM in response to hypoxia (Phillipson et al. 1978; Henderson-Smart & Read, 1979; Jeffery & Read, 1980; Neubauer et al. 1981; Fewell & Baker, 1987) and airway stimulation or obstruction (Sullivan et al. 1978; Davidson & Fewell, 1994).

In contrast to animal studies, investigations in humans have not found hypercapnic arousal to be depressed in REM. One study in adult humans found arousal to hypercapnia occurred at the same time in both REM and non-REM sleep states (Douglas et al. 1982) and another found a brisker, though transient (2–4 s) arousal in REM sleep (Berthon-Jones & Sullivan, 1984). In the human newborn arousal responses to hypoxia and hypercapnic hypoxia are enhanced, not depressed, in active sleep compared with quiet sleep (Campbell et al. 1998; Parslow et al. 2003). Further comparative studies are needed to establish the underlying basis for the reversed sleep state-specific arousal differences in sleeping humans compared with animals. Other human studies will be needed to establish whether arousal is impaired with repeated episodes of hypoxia or hypercapnic hypoxia.

Cardio-respiratory responses to repetitive hypercapnia

When examined over the entire 40 exposures to hypercapnia, ventilation increased significantly during hypercapnia in both sleep states, but this response was significantly reduced in active sleep compared with quiet sleep (P < 0.05). Previous studies examining the ventilatory response to hypercapnia are conflicting, with some studies reporting a reduced response in active sleep compared with quiet sleep (Douglas et al. 1982; Honma et al. 1984; Moriette et al. 1985; Cohen et al. 1991; Praud et al. 1991), while others have reported no difference between states (Guthrie et al. 1980; Haddad et al. 1980; Rigatto et al. 1980; Kalapesi et al. 1981). REM sleep has tonic and phasic components, and consideration of these phases in adult dogs (Phillipson et al. 1977) reveals that the ventilatory responses to hypercapnia may be lower in phasic REM, but not tonic REM, in comparison with quiet sleep (Sullivan et al. 1979). In the lamb, as in the newborn of other species (Roffwarg et al. 1966), active sleep is predominately phasic in nature, which is in keeping with the lower ventilatory response in active sleep overall.

By employing endotracheal CO2 administration, our preparation by-passed laryngeal CO2 receptors that reflexly inhibit ventilation (Wang et al. 1999). Thus, the ventilatory responses that we observed may have been less had the entire airway been exposed to CO2. However, the extent of any such depression is likely to be very small as the laryngeal chemoreflex is relatively minor compared with the carotid chemoreflex in its influence on ventilation. Moreover, the presence of the laryngeal chemoreflex in young animals is uncertain; in young piglets, 10% intralaryngeal CO2 had no effect on ventilation (Heman-Ackah & Goding, 2000). In any event, as neither ventilation nor arousal were substantially altered by repeated CO2 exposures, it is not likely that our findings would be affected by including the laryngeal airway in our tests.

The circulatory responses to hypercapnia have been less well examined. The response to inhalation of small amounts of CO2 during wakefulness in the adult human results in increased cardiac output and heart rate, and peripheral vasodilatation (Bristow et al. 1971). The characteristic response in awake lambs is a re-distribution of cardiac output, so that blood flow to the brain is increased and blood flow to the periphery is decreased (Rosenberg et al. 1984). The effects of hypercapnia on heart rate and blood pressure responses in the newborn remain largely unexplored, and data on the usual cardiovascular response to hypercapnia in sleep are few (Fewell & Baker, 1989). In this study, there was a very small but significant increase in blood pressure and decrease in heart rate in response to hypercapnia during quiet sleep, while cardiovascular changes did not reach significance in active sleep. A previous study in lambs found that there were no significant changes in heart rate or blood pressure during hypercapnia in either sleep state (Fewell & Baker, 1989). Thus, although clarification on the usual cardiovascular responses are required, data from this study support the idea that the ventilatory and cardiovascular responses to hypercapnia are poorer in active sleep than quiet sleep.

Although the overall cardio-respiratory responses to hypercapnia may be reduced in active sleep compared with quiet sleep, the responses persist with repetition of hypercapnia, similar to the arousal responses. This is in marked contrast with the responses to hypoxia, in which there is a rapid, concurrent abolition of the arousal, ventilatory and blood pressure responses in active sleep after just 10 exposures to hypoxia (Johnston et al. 1998, 1999). Moreover, like hypoxia, arousal deficits rapidly emerge after successive exposures to hypercapnic hypoxia (Waters & Tinworth, 2005). It appears therefore that it is the response to hypoxia per se that is depressed with repetition. Hypoxic depression of ventilatory responses in these studies of sleeping animals is in contrast to the ventilatory stimulation known as long-term facilitation (LTF) that outlasts the stimulus of episodic hypoxia (Mitchell & Johnson, 2003). LTF is particularly prominent in anaesthetized and decerebrate animals, but is less well expressed and sometimes absent in awake animals and humans (Powell et al. 1998). As several opposing facilitatory and inhibitory processes modulate breathing (Mitchell & Johnson, 2003), absence of LTF in REM sleep may represent an example of dominance of a depressive process in this behavioural state.

Data from this study add to the concept that sleep state has a profound influence on protective responses to asphyxial stress. Ventilatory responsiveness is reduced in active sleep in hypercapnia (Cohen et al. 1991; this study), hypoxia (Henderson-Smart & Read, 1979; Douglas, 1985) and asphyxia (Campbell et al. 1998). Though there are no progressive changes with repeated hypercapnia, arousal and ventilatory responses are both subject to depression with repetition of hypoxia in lambs in active sleep, and ventilation is further impaired when arousal does not occur (Johnston et al. 1998, 1999).

In contrast to the species differences in arousal sensitivity, ventilatory responses to asphyxial stimuli appear to be depressed in REM and active sleep in humans just as in animals (Douglas, 1985; Campbell et al. 1998). Thus, the ability of the newborn to instigate protective cardio-respiratory responses appears to be more robust in quiet sleep, and subject to impairment in active sleep, rendering the latter state more vulnerable to asphyxial stress.

In summary, arousal and cardiorespiratory responses to hypercapnia are both less effective in active sleep compared with quiet sleep. Unlike hypoxia and asphyxia, repeated exposure to hypercapnia does not lead to a depression of responses. Nevertheless, the poorer responses to hypercapnia that characterize active sleep will add to the vulnerability to asphyxia in this sleep state.

Appendix

Acknowledgements

This work was supported by the National Health and Medical Research Council of Australia (A.W., D.G.) and the Sudden Infant Death Research Foundation of Victoria, Australia (R.J.).

Ancillary

Advertisement